The Effectiveness of Halting Measures in Constraining North Korea’s Nuclear Weapons Program

ABSTRACT

This paper explores the concept of ‘irreversibility’ in relation to denuclearization efforts on the Korean Peninsula over the past 30 years, covering the most impactful US, South Korean, and North Korean approaches during this timeframe. Attempts to constrain North Korea’s nuclear programme have largely been characterized by ‘halting measures’, such as technical freezes or political moratoria; however, these measures did not have lasting effects. This paper examines the systemic barriers that challenged the sustainability of these halting measures, in particular disagreements over goals, terminology, sequencing, as well as a failure to achieve sustained momentum during challenging periods of domestic politics. It then considers what this suggests about the viability of these types of measures in promoting ‘irreversible’ disarmament in North Korea moving forward.

KEYWORDS:

• North Korea
• nuclear weapons
• irreversibility
• disarmament
• missile program

The question of ‘irreversibility’ in arms control, non-proliferation, and disarmament has evolved significantly since it came to enhanced prominence in the 1990s. Initially seen as a zero-sum objective—wherein a state’s nuclear dismantlement would be considered either reversible or irreversible––in recent years experts have aimed to frame irreversibility as a spectrum (Anthony 2011; Cliff and Persbo 2011; International Partnership for Nuclear Disarmament Verification 2018; Ogilvie-White et al. 2023). Notably, recent scholarship has highlighted how irreversibility, despite being widely used in multilateral arms control and disarmament, is seldom used in a consistent way. One proposed approach is to define irreversibility as ‘a feature or quality of a disarmament or an arms control process that involves limiting the capacity for re-armament including the possible reconstitution of aspects of weapons programmes’ (Elbahtimy 2023).

In this context, the lowest degree of ‘irreversible’ disarmament could be satisfied even if the entire nuclear production complex remains intact solely by the dismantlement of nuclear warheads, while the highest degree would also include abandoning the entire means of production. As such, ‘adequate irreversibility’ of disarmament measures could be achieved by enacting a comprehensive patchwork of political, legal, normative, and technical actions that would be highly challenging and expensive for a state to overturn (Ritchie 2023). A combination of these actions over time could lead to disablement, and, if these steps are robust enough so that they cannot be readily undone or resumed, eventually towards a higher degree of ‘irreversibility’ through a decreased capacity to reverse decisions and reestablish processes to rebuild nuclear weapons and related infrastructure.

The North Korean case presents a relatively unique challenge for this irreversibility framework. While several other well-documented country cases––including South Africa, Kazakhstan, Belarus, and Ukraine––successfully resulted in comprehensive nuclear dismantlement (Kassenova 2024), the past three decades of North Korean disarmament measures have largely been characterized by technical freezes or political moratoria that did not have lasting effects. If freezes and moratoria primarily involve halting a country’s progress towards nuclear proliferation, rather than reversing it, then this raises legitimate questions as to where these types of halting measures fall on the spectrum of ‘irreversible’ nuclear disarmament (Elbahtimy and Peel 2023).

To that end, this paper examines the link between irreversibility and negotiations with the DPRK over its nuclear programme. It particularly examines the role of freezes and moratoria as tools to constrain DPRK’s nuclear programme. It also considers how these halting measures were unsuccessful in the North Korean context, and what this suggests about the viability of these types of measures in promoting ‘irreversible’ disarmament moving forward.

This paper starts by examining the various negotiated freezes and moratoria applied to DPRK’s nuclear programme. It specifically examines measures applied at Punggye-ri nuclear test site, liquid-engine test stand at Sohae, and the Yongbyon complex as key examples of facilities where such halting measures were applied. It then examines the systemic barriers that challenged the sustainability of these halting measures, and considers the factors that undermined the abilities of negotiators to move beyond freezes and moratoria in constraining North Korea’s nuclear programme. Disagreements over goals, terminology, sequencing, as well as a failure to achieve sustained momentum during challenging periods of domestic politics, hindered progress on achieving sustainable irreversibility measures for North Korea’s nuclear program. Ultimately, until some of these barriers are reduced, future halting measures are unlikely to yield sufficient progress to achieve ‘adequate irreversibility’.

Freezes and Moratoria

The halting measures employed in the context of nuclear negotiations with North Korea can be roughly grouped into two categories which, for the purposes of this paper, are labelled ‘technical freezes’ and ‘political moratoria’. In the North Korean context, freezes have comprised disablement actions that sought to halt ongoing technical progress on a specific process, such as plutonium production. Moratoria, on the other hand, have comprised political declarations to abstain from certain high-profile activities, such as long-range missile launches or nuclear weapons tests, for a period of time. A key distinction between the two is that while technical freezes have traditionally resulted in operational work stoppages, political moratoria have generally not precluded substantial technical progress from taking place outside of the specific action that was being halted.

Both types of halting measures can act as stepping stones along the path towards eventual disarmament, and in some cases could be considered necessary precursors to more ‘irreversible’ measures down the line (Hecker, Carlin, and Serbin 2019). It is difficult to imagine dismantling a plutonium reprocessing facility, for example, without first halting the reprocessing of plutonium altogether.

In an ideal world, these halting measures would be followed by measures designed to roll back and ultimately eliminate key elements of a nuclear weapons program. Under certain circumstances, however, the halting measures can themselves be highly consequential steps in pursuit of ‘adequate irreversibility’. This is particularly true for technical freezes of certain processes which, if frozen for long enough, would be highly challenging to resume. For example, while the United States is legally required to maintain the capability to perform an underground nuclear test within 24 to 36 months if deemed necessary, scholars from the Johns Hopkins Applied Physics Laboratory concluded in 2022 that in the 30 years since the United States last conducted an underground nuclear test, ‘the test teams have long since dispersed and the associated firsthand knowledge base has atrophied. Moreover, most of the equipment, facilities, and supporting infrastructure have long since fallen into disuse and would have to be reconstituted’––thus making these timelines potentially unrealistic (Frankel, Scouras, and Ullrich 2022). In another example, SIPRI researcher Robert E. Kelley has proposed a ‘tritium cut-off-treaty’ with the goal of using tritium’s short half-life of 12.3 years to ‘starve’ nuclear-armed countries of the critical material for boosting the yields of their warheads (Kelley 2020). These examples suggest that in certain circumstances, as time passes the effectiveness of halting measures in pursuing ‘adequate irreversibility’ can be strengthened.

A review of the past three decades of North Korean nuclear negotiations suggests that technical freezes––even if they were ultimately reversed––were successful in stalling certain elements of North Korea’s nuclear weapons program. However, given that these halting measures were largely instituted in isolation from one another and did not last for prolonged periods of time, North Korea ultimately did not face significant roadblocks to resuming and reconstituting any activities that had previously been paused. To explore this in greater depth, three notable case studies will be examined below.

Throughout the past three decades of North Korean nuclear negotiations, Pyongyang enacted significant dismantlement measures at three distinct sites: the Punggye-ri nuclear test site, the Sohae launch complex, and at Yongbyon. It is clear, however, that nearly every step that was taken by North Korea could ultimately be reversed relatively quickly––some activities within even just a few days. While these steps were ultimately meant to build on each other, the relative degree of rapid reversibility for each measures precluded sustained progress and served to undermined confidence in North Korea’s genuine interest in disarmament.

Punggye-Ri Nuclear Test Site

During Kim Jong Un’s April 2018 speech where he announced a moratorium on nuclear weapons and long-range missile tests, he also stated that ‘the northern nuclear testing ground [Punggye-ri] will be dismantled to transparently guarantee the discontinuation of nuclear test’, a move which he called ‘an important process for global nuclear disarmament’ (Pyongyang Times 2018).

In May 2018, approximately two dozen international journalists were invited to witness the destruction of three portals and several buildings at the Punggye-ri nuclear testing site. The observers were not allowed to bring any monitoring equipment with them, and no non-proliferation experts were invited to witness the detonations. North Korean officials reportedly provided very little technical information about the procedure and thus ‘it was unclear [to those present] whether the explosions rendered the tunnels inoperable, or only caused limited damage’ (Ripley, Schwarz, and Devitt 2018). Additionally, satellite imagery confirmed the presence of ongoing maintenance and activity at the site, indicating that it had not been fully abandoned (Bermudez and Cha 2019b). Given these lingering doubts, both South Korea and the United States publicly claimed that the destruction at Punggye-ri was only a suspension and could be reversed (G. Kim 2019; US Department of State 2019).

These predictions of reversibility ultimately came to fruition: in September 2022, the UN Panel of Experts on North Korea noted that the country had ‘reopened its nuclear test infrastructure, including the test tunnel and supporting buildings which were demolished in May 2018’ (United Nations Security Council 2022, 6–9). The Panel of Experts observed vehicle movements, intensive construction, cable installation, and re-excavation activities consistent with the reactivation of the Punggye-ri site. The Panel of Experts further noted that ‘[a]s of early June [2022], two Member States assessed that the preparation for nuclear tests was at a final stage’ (United Nations Security Council 2022, 6–9). Although a seventh nuclear test had not occurred by the beginning of 2024, it seems apparent that the prospect of North Korea conducting such a test is no longer a matter of ‘if’, but ‘when’.

Ultimately, the steps that North Korea took to ‘dismantle’ the Punggye-ri nuclear test site were largely cosmetic and proved to be completely reversible. As Ankit Panda noted in his book Kim Jong Un and the Bomb, ‘The May 2018 demolition of the Punggye-ri nuclear test site was presented to the world as a major gesture of willingness to pursue “denuclearization”, but its reversibility demonstrated how careful Kim was being about what steps he was willing to take as the diplomatic process kicked off’ (Panda 2020, 269).

Sohae Liquid-Engine Test Stand

During the Singapore Summit in June 2018, Kim Jong Un committed to dismantling the missile engine test site at Sohae (Spetalnick 2018). Sohae was an important symbolic and programmatic site for North Korea’s missile program, as it was the key location where Kim Jong Un had validated the liquid-fueled engines for North Korea’s longer-range missiles (LaFoy 2018).

Throughout July and early August 2018, satellite imagery analysts were able to track the partial dismantlement of the rail-mounted processing building––used for final check-out of the rocket before launch––as well as the rocket engine test stand and the bunkers storing fuel and oxidizer (Bermudez 2018a, 2018b). When US-DPRK talks broke down in August, however, the activities at Sohae ground to a halt as well (Bermudez, Pabian, and Liu 2018). The UN Panel of Experts subsequently concluded that the initial dismantlement activities ‘may be reversed as only light or moveable parts, such as the metal superstructure, were dismantled’ (United Nations Security Council 2019, 356).

After negotiations more permanently broke down after the second Trump-Kim summit in February 2019, Kim Jong Un immediately began to reestablish Sohae as an operational test site. Only two days after the delegations departed Hanoi, satellite imagery indicated that North Korea had begun to rebuild the vertical engine test stand and reposition the rail-mounted transfer facility to its original location (Bermudez and Cha 2019a, 2019c, 2019d). Sohae was reconstructed within roughly one week, confirming the prior assessment of the UN Panel of Experts that North Korea’s initial dismantlement efforts were easily reversible (Bermudez and Cha 2019d).

Yongbyon Complex

As the ‘beating heart’ of North Korea’s nuclear program, Yongbyon has been the primary focus for every major disarmament effort with North Korea, and at several distinct occasions over the past three decades, North Korea has taken steps to physically disable the complex (Panda 2020, 276). However, despite the successes in freezing activities at Yongbyon, at each juncture when negotiations ultimately collapsed, North Korea was able to revitalize the complex very quickly.

For example, after the Agreed Framework collapsed in late 2002, North Korea was able to remove all IAEA monitoring equipment in a three-day period, move 1,000 previously canned fuel rods back into the reprocessing facility in another three-day period, resume ‘normal’ reactor operations one month later, and complete its reprocessing campaign by the end of June 2003 (four months later)––resulting in an estimated 25 to 30 kilograms of plutonium (Brooke 2002; Hecker 2023b; Hecker, Carlin, and Serbin 2018a; Pollack 2003; Struck 2003).

In subsequent years, multiple rounds of Six-Party Talks yielded diluted proposals that closely mimicked what had previously been included in the Agreed Framework––all with Yongbyon at the center of a potential agreement. The sixth round of talks in February 2007, however, culminated in a breakthrough: The United States and North Korea agreed to a list of 12 ‘disablement actions’ that would be undertaken at three different Yongbyon facilities: the fuel fabrication facility, the reactor, and the reprocessing facility (Hecker 2008, 3–4). Notably, the US list differed slightly from the list presented by Yongbyon officials in that it did not include step four, it combined steps five and six, and apparently included one additional item: the disablement of fresh, unclad fuel rods fabricated prior to 1994 (Hecker 2008, 3). According to Dr. Hecker, ‘the operational definition of “disablement” [was] to make it more difficult, but not impossible, to restart the nuclear facilities’(Hecker 2008, 3–4). This suggests a degree of nuance from the United States in understanding that the goal of pursuing ‘irreversibility’ was not to prevent North Korea from ever restarting its nuclear program (presumably an unattainable goal), but rather to make it much more difficult to do so. In a statement before the Senate Committee on Armed Services, Assistant Secretary for East Asian and Pacific Affairs Chris Hill confirmed that, ‘Upon completion of all 11 steps, the DPRK would have to expend significant effort, and time – upwards of 12 months – to reconstitute all of the disabled facilities’ (US Department of State 2008).

The list of proposed ‘disablement actions’ was as follows:

Fuel Fabrication Facility:

1. Removal and storage of all three uranium ore concentrate dissolver tanks;

2. Removal and storage of all seven uranium conversion furnaces, including storage of refractory bricks and mortar sand;

3. Removal and storage of both metal casting furnaces and vacuum system, and removal and storage of either machining lathes;

4. Storage of the remaining UO3 powder in bags with monitoring by IAEA (this constitutes nearly five tons of powder);

Five Megawatt Reactor:

1. Cut and removal of portions of steel piping of the secondary cooling loop outside the reactor building;

2. Removal of the wood interior structure of the cooling tower;

3. Discharge of 8000 spent fuel rods;

4. Removal and storage of the control rod drive mechanisms;

Reprocessing Facility:

1. Cutting of cable and removal of drive mechanism for trolley that moves spent fuel caskets from the fuel receiving building into the reprocessing facility.

2. Cutting of two out of the four steam lines into the reprocessing facility;

3. Removal of the crane and door actuators that permit spent fuel rods to enter the reprocessing facility (at Level −1);

4. Removal of the driving mechanisms for the fuel cladding shearing and slitting machines (at Level −1).

When the disarmament process ultimately began to fail in 2008 and 2009, North Korea was able to restart Yongbyon at a similar pace to how it had done so after the collapse of the Agreed Framework. Only ten days after the IAEA’s expulsion from the country in April 2009, North Korea announced that it had restarted the reprocessing facility at Yongbyon (Hecker 2009). North Korea subsequently declared that it had completed its reprocessing campaign by the end of August 2009––a similar four-month pace to its previous campaign (Nikitin 2013, 20). The reactor itself remained mothballed until April 2013, when North Korea announced that it would restart the facility. Satellite imagery suggested that the reactor subsequently began operating within six months (Eckert and Mohammed 2013; Harlin 2013).

For the next five years, no dismantlement activity took place at Yongbyon until the meeting between Moon Jae-in and Kim Jong Un in September 2018, when North Korea agreed, once again, to ‘permanently dismantle’ the Yongbyon site (National Committee on North Korea 2018). In December 2018, North Korea shut down its reactor as a confidence-building measure in advance of the planned February 2019 Trump-Kim summit in Hanoi; however, after the summit ended in failure, the reactor restarted operations in July 2021 (Board of Governors General Conference 2021, 4; Reuters 2019).

The disablement steps that North Korea took throughout the past three decades represented a proof-of-concept for how halting measures could be utilized to make it more difficult for North Korea to restart nuclear operations after a prolonged pause. However, given that these measures were only ever adopted in relative isolation, and did not last long enough to meaningfully challenge North Korea’s ability to reconstitute, they ultimately did not succeed in paving the way for ‘adequate irreversibility’.

Systemic Barriers

Why did negotiations with North Korea not succeed in moving beyond initial halting measures like freezes or moratoria? A review of the past three decades of engagement with North Korea suggests that the United States and South Korea at times largely viewed irreversibility through a zero-sum lens, thus limiting progress on more actionable and sustainable North Korean disarmament measures that would lead to ‘adequate irreversibility’. Ultimately, disagreements over goals, terminology, and sequencing, as well as a failure to achieve sustained momentum during challenging periods of domestic politics, hindered progress on achieving sustainable irreversibility measures for North Korea’s nuclear program.

Competing Priorities, Shifting Goalposts

Despite the dramatic evolution in North Korea’s nuclear capabilities over the past three decades, the United States has largely framed the question of a nuclear North Korea in the same way: as a proliferation problem necessitating a non-proliferation solution. This approach initially proved constructive during the first joint efforts at denuclearization, as North Korea’s nuclear program was still in its infancy and Pyongyang appeared to demonstrate a cautious willingness to dismantle its nascent nuclear program in exchange for economic and security incentives from the United States.

The first major denuclearization deal between the United States and DPRK––the 1994 Agreed Framework––called for North Korea to freeze the operation of its 5 MW(e) reactor and plutonium reprocessing plant at Yongbyon, as well as freeze construction of two other reactors under construction (US Department of State 1994).

As a nonproliferation strategy, the Agreed Framework was successful in advancing US objectives: not only was plutonium production at Yongbyon paused for nearly a decade as a result of the freeze under the Agreed Framework, but North Korea completed canning and sealing the 8,000 spent fuel rods from the 5 MW(e) reactor––containing enough plutonium to produce four to five nuclear weapons––and halted construction of two other reactors, which could have allowed North Korea to produce up to 300 kg of plutonium annually (Hecker 2023a, 83).

Although the Agreed Framework emphasized an immediate freeze on North Korea’s nuclear activities, the eventual goal––as spelled out in the agreement and reaffirmed by both parties––was for North Korea to ‘eventually dismantle these reactors and related facilities’ (US Department of State 1994). A freeze was meant to be the first major step before further measures were taken. Ending the freeze at Yongbyon was believed to be the “surest and quickest path for North Korea to obtain nuclear weapons” (US Department of State 1999).

The Agreed Framework did not include any language referencing ‘irreversibility’ as a declared objective of the deal, but this may have been an intentional choice on the part of the US negotiating team. Robert Gallucci, the Clinton administration’s chief negotiator with North Korea during the Agreed Framework, noted his concerns over how the United States’ pursuit of irreversibility had evolved since the 1990s: ‘My concern here is that we latch onto a standard which is, physically, not actually plausible’, Gallucci told NPR in 2018. ‘There’s no way of doing something that’s irreversible, that I know of’ […] ‘As long as the scientists and engineers who designed and built the facilities are still alive and living in the country, it can be built again’ (Welna 2018).

Following the collapse of the Agreed Framework in 2003 and the arrival of the Bush administration, the United States gradually shifted its requirements for North Korean concessions, driven by a ‘perfection over progress’ approach that emphasized ‘irreversible’ dismantlement. The Bush administration and Congressional Republicans criticized the Agreed Framework for not going far enough to dismantle North Korea’s nuclear program, suggesting that Washington was rewarding North Korea’s belligerency ‘in exchange for promises’ (Dole 1994; Lewis 2015). The Bush administration even started using the word ‘disablement’ more frequently to distance itself from the rhetoric of the prior administration that they had so harshly criticized (Lewis 2015).

The Bush administration’s diplomatic efforts centered around the ‘Six-Party Talks and worked with Japan and South Korea to formalize ‘irreversibility’ within the discussions under the umbrella of ‘CVID’––’complete, verifiable, and irreversible denuclearization’ (Republic of Korea Ministry of Foreign Affairs 2003). Washington in particular insisted that North Korea agree to these terms before it would provide any concessions or security guarantees, and US negotiator James Kelly was instructed to stick to this formula in an effort to duplicate a ‘dismantlement-up-front’ strategy that had recently proved successful in Libya (Hecker 2023a, 43, 90).

During this time, key political figures continued to insist on a hardline stance against North Korea. In 2004, the Bush administration outlined its approach to dealing with North Korea’s nuclear program as follows:

we seek the complete, verifiable, and irreversible dismantlement of its nuclear programs, nothing less. We cannot accept another partial solution that does not deal with the entirety of the problem, allowing North Korea to threaten others continually with the revival of its nuclear program (Kelly 2004a, 2004b).

North Korea, for its part, labelled the acronym ‘disgusting’, and threatened to stop engaging with the United States if it insisted on pursuing CVID (Washington Times 2004). When the fourth round of six-party talks took place in September 2005, participants concluded a landmark joint statement of principles to guide future negotiations, in which North Korea committed ‘to abandoning all nuclear weapons and existing nuclear programs’ and returning to the NPT and IAEA safeguards. However, Washington hardliners crafted a US unilateral statement based on its interpretation of the agreement, insisting that ‘[a]ll elements of the DPRK’s past and present nuclear programs––plutonium and uranium––and all nuclear weapons will be comprehensively declared and completely, verifiably, and irreversibly eliminated’ (Hecker 2023a, 122). This came after North Korea had already rejected CVID language in past negotiations, and after the 2005 Joint Declaration had been carefully designed to not include this phrasing.

The final round of six-party talks culminated in a breakthrough 2007 agreement to complete the disablement of the 5 MW(e) reactor, reprocessing plant, and fuel fabrication plant at Yongbyon (US Department of State 2007a, 2007b). The disablement process began to derail, however, over North Korea’s disagreement with the widening scope of the US proposed verification measures, which insisted upon ‘full access to any site, facility or location’ deemed relevant to the nuclear program (Aoki 2018, 139; Kessler 2008). According to reporting at the time, US negotiators had opposed the far-reaching verification plan as a non-starter but had been overruled by higher-level officials (Kessler 2008).

After North Korea’s nuclear program reached a level of maturity under Kim Jong Un, the United States was left with few prospects for pursuing CVID. Nonetheless, in the summer of 2018, Secretary of State Mike Pompeo introduced a re-brand of this objective with a new phrase––’final, fully verified denuclearization of North Korea’, or ‘FFVD’––which he erroneously claimed that Kim Jong Un had agreed to during the recent 2018 Singapore Summit (Pompeo 2018a, 2018b).

When Trump administration officials finally clarified what exactly FFVD would entail, it appeared to mirror key clauses found in the 2005 Joint Statement of the Six-Party Talks (A. Kim 2019). The context for these negotiations, however, had shifted dramatically: by this time, Kim Jong Un possessed a maturing nuclear arsenal and long-range missiles that could potentially deliver warheads to intercontinental ranges. Essentially, North Korea had shifted from a proliferation problem to a deterrence problem, but the US position had not shifted with it.

The discrepancy was made publicly clear when President Trump provided Kim Jong Un with a list of demands during the 2019 Hanoi Summit. The document, believed to have been strongly influenced by National Security Advisor John Bolton, insisted that Kim Jong Un acquiesce to the FFVD conditions, as well as ‘fully dismantl[e] North Korea’s nuclear infrastructure, chemical and biological warfare program and related dual-use capabilities; and ballistic missiles, launchers, and associated facilities’ and transfer all of North Korea’s nuclear weapons and fissile material to the United States’ (Wroughton and Brunnstrom 2019). This effectively ended the summit before it began, and reportedly left Kim Jong Un feeling furious and humiliated (Panda 2020, 277).

Three years later, in September 2022, North Korea passed a new law codifying North Korea’s nuclear doctrine, which––in a direct rebuttal to CVID––Kim Jong Un said made his country’s nuclear status ‘irreversible’ (Rodong Sinmun 2022). Such announcements––coupled with sophisticated technological advancements in both its nuclear weapons and missile programs––suggest that North Korea is eager to convince the United States to abandon CVID entirely. However, as of the beginning of 2024, the Biden administration has largely continued to abide by a CVID approach that has proven unsuccessful in achieving North Korean disarmament. South Korean President Yoon Suk-yeol has also resurrected the ‘CVID’ language in his public statements––a stark deviation from those of his predecessor, who conspicuously eschewed nods to both verification and irreversibility (Lee 2022). Given that US policy continues to emphasize irreversible denuclearization over actions that could advance ‘adequate irreversibility’, the two sides are likely to remain deadlocked (Brunnstrom and Lewis 2022).

Sequencing Stalemate

The contrasting US and North Korean viewpoints on sequencing have left little room for compromise: North Korea has preferred synchronous, step-by-step measures taken by both parties, whereas the United States has insisted that North Korea fulfill the primary elements of CVID/FFVD before making any concessions, such as sanctions relief (A. Kim 2019).

The proposed compromise was enshrined during the Six-Party Talks, wherein all parties agreed to implement disarmament steps in exchange for economic assistance under the principle of ‘commitment for commitment, action for action’ (US Department of State 2005). As the Bush administration explained to Congress, however, it was incumbent upon North Korea to take the first step: ‘Under the framework, DPRK receives something from the other parties only as it moves forward in its own commitments’ (Hill 2008).

Despite this compromise, at several crucial junctures during negotiations, both sides failed to hold up their ends of the bargain. Throughout the Agreed Framework implementation period, the United States and its partners failed to deliver their heavy fuel oil and light-water reactor commitments on-time due to interoperability challenges, a lack of funding, growing Congressional resistance, and North Korea’s own intractability (Aoki 2017; Senate Committee on Foreign Relations 1998). Consequently, Pyongyang maintained that it would not complete dismantlement of its reactors and related facilities until after the United States had followed through on the construction of both light-water reactors as agreed to in years prior (Hecker 2023a, 83). However, the Bush administration’s insistence upon ‘irreversible’ dismantlement, as well as refusing to ‘reward North Korea for coming into compliance with its past obligations’, served to undermine Congressional support for providing North Korea with those same reactors (Hecker 2023a, 94; Hill 2005). This created a sequencing problem, wherein both sides expected the other to fulfil its side of the deal first.

This framing outlived the Bush administration: Andrew Kim, who helped orchestrate the 2018 Singapore Summit, noted that ‘lessons of the past place the burden of proof on the North’ to move first (A. Kim 2019). Meanwhile, North Korea has come to the opposite conclusion, that in 2018 they took several grand gestures towards denuclearization and were frustrated that the United States did not reciprocate.

Discrepancies in Terminology

One of the most consequential obstacles to tangible progress on North Korea’s nuclear dismantlement has been noticeable disagreements over the terminology used to describe various objectives. These disagreements have ranged from technical language to overarching goals.

One particularly high-profile example of a misunderstanding over terminology prompted the collapse of the 2012 Leap Day Deal, a ‘food-for-freeze’ arrangement that called for a moratorium on long-range missile launches, nuclear tests, and certain fissile material production at Yongbyon, as well as the readmittance of IAEA inspections who had been expelled in 2009 (Panda 2020, 65).

Despite Obama administration officials calling the arrangement the ‘Leap Day Deal’, in reality the settlement was far from a negotiated ‘deal’. Instead, both the United States and North Korea released unilateral statements in February 2012 essentially summarizing the content of high-level talks the previous year, and despite US insistence at the time that the two countries were in alignment, the statements contained glaring discrepancies (Lewis 2012). Notably, the two statements differed on whether reprocessing activities would be included in the moratorium, and the North Korean statement explicitly did not include a commitment to ‘disable’ the 5 MW(e) reactor at Yongbyon, as written in the US statement (KCNA 2012; Nuland 2012).

As notable as those differences were, however, it was a misunderstanding over a different phrase included in both statements––‘long-range missile launches’––that ultimately collapsed the ‘deal’. Given the technological similarities between space-launch vehicles and long-range missiles, the United States had apparently warned North Korea that ‘any missile test, for any purpose, would violate the terms of the agreement under negotiation … ‘ (Lewis 2012). North Korea, for its part, had long distinguished between the two types of launches and explicitly stated in March that ‘the DPRK had already consistently clarified at the three rounds of the DPRK-US high-level talks that [a] satellite launch was not included in the long-range missile launch moratorium’ (Hecker 2023a, 257). When Kim Jong Un ultimately launched a satellite in April 2012 from Sohae to commemorate the centenary of former leader Kim Il Sung, the Obama administration suspended the Leap Day ‘deal’.

A less consequential, but highly illuminating, example of unclear terminology can be seen in the Pyongyang Joint Declaration made by President Moon and Chairman Kim in September 2018. The agreement specifically conceded North Korea’s Sohae complex, noting that it would be dismantled under the observation of international experts (National Committee on North Korea 2018). However, the question of irreversibility played a key role in triggering confusion between what had actually been agreed between the two sides. While the two sides agreed to a unitary Korean-language text, their English translations contained crucial terminological differences: the South Korean version stated that North Korea would ‘permanently dismantle’ Sohae, whereas the North Korean version agreed to ‘permanently shut down’ the site. The former clearly portrayed an intention to physically tear down the site, whereas the latter only committed to a freeze of operations (Panda 2020, 270–271).

Perhaps most importantly, the phrase ‘denuclearization of the Korean Peninsula’ has been used by both North and South Korea since the 1992 Joint Declaration, but has been frequently misunderstood––or, at times, deliberately misconstrued––by US interlocutors (United 1992). While North Korea has interpreted the phrase to include the elimination of US extended deterrence commitments to South Korea, officials from both the Trump and Biden administrations have instead interpreted it as a North Korean offer to disarm––even going so far as to use the phrase ‘denuclearization of North Korea’––which has actively inhibited progress on negotiations (KCNA 2018; Nichols 2021; The White House 2021).

Failure to Take Advantage of Momentum

The concept of ‘momentum’ as it relates to disarmament has been relatively underexplored but appears to have had a significant impact on how policymakers pursued irreversibility throughout negotiations with North Korea. Some policymakers viewed initial physical momentum kickstarted by technical dismantlement measures as critical for achieving a high degree of irreversibility in the long term. In addition, political momentum, prioritized primarily by the more conciliatory left flank of officials in South Korea, was also seen as necessary for setting the stage and broadening the aperture for further denuclearization efforts.

Physical Momentum

Following his first summit with Kim Jong Un in June 2018, President Trump announced his intention to achieve North Korean denuclearization as fast as possible; he rebutted suggestions that it could take up to 15 years and claimed that ‘there will be a point at which when you are 20% through you can’t go back’ (Trump 2018). Trump’s analysis was not necessarily incorrect; while completely eliminating every individual component of North Korea’s nuclear program could take many years, several expert analysts noted at the time that front-loading the most critical activities––in particular, halting nuclear testing, fissile material production, and long-range missile testing––could naturally lead to a gradual rollback of North Korea’s nuclear program that would be highly challenging to reverse (Hecker, Carlin, and Serbin 2018b; Kelley 2018).

In particular, halting the production of fissile materials for an extended period of time had long been seen as a potential boon for irreversibility. After a visit to Yongbyon in February 2008, Dr. Siegfried Hecker noted that a freeze––absent regular maintenance––constituted an irreversibility measure in and of itself:

When asked how long they can do without maintenance and still be able to salvage the facilities, [Yongbyon officials] said that the ability to restart the facility vanishes if maintenance restrictions last for a long time (they did not define what they mean by long) (Hecker 2008, 6).

While the disablement process ultimately failed when North Korea walked away from the Six-Party Talks in April 2009, Dr. Hecker estimated that the delay likely cost North Korea between six to 18 months in order to ‘regain their prior production rate of six kilograms (or roughly one bomb’s worth)’ (Hecker 2008, 1; KCNA 2009). Although these delays were beneficial in and of themselves, the fits and starts that characterized nuclear negotiations with North Korea––wherein North Korea would repeatedly freeze and subsequently unfreeze its nuclear facilities after talks failed––inhibited the momentum that would have been needed to yield more irreversible benefits.

Political Momentum

South Korean policymakers, particularly the Sunshine-oriented Moon Jae-in administration, emphasized pursuing opportunities for political momentum to set the stage for technical steps towards denuclearization. In particular, throughout his administration, President Moon explicitly tied the legal status of the Korean War to the broader issue of North Korea’s nuclear program.

Some analysts have noted that ending the Korean War would be little more than a symbolic gesture, and that it would do little to reduce military tensions on the Korean Peninsula (Klingner 2020). Throughout his term, however, President Moon insisted that bringing a formal end to the war would necessarily set the stage for ‘irreversible progress’ on denuclearization. In particular, after denuclearization talks with North Korea stalled in 2019, Moon made several public statements claiming that ‘[t]he end-of-war declaration will, indeed, open the door to complete denuclearization and permanent peace regime on the Korean Peninsula’ (Je-Hun 2021). He reaffirmed this sentiment in his UN General Assembly speech the following year: ‘When the parties involved in the Korean War stand together and proclaim an end to the war, I believe we can make irreversible progress in denuclearization and usher in an era of complete peace’ (Je-Hun 2021).

These statements illuminate how important President Moon considered momentum to be in laying the groundwork for ‘irreversible’ progress on denuclearization. In support of Moon’s conclusions, recent scholarship has suggested that for several other WMD-related agreements outside of the Korean context, political factors have often been more successful in setting the stage for sustained progress on irreversible disarmament (Rodgers and Williams 2023).

Conclusion

Today, the United States and South Korea are far from achieving what many policymakers would consider ‘adequate irreversibility’ of North Korea’s nuclear weapons program. In the decades that have passed since North Korea first considered denuclearization, the country has developed an increasingly sophisticated arsenal and adopted policies that has further ingrained its reliance on its nuclear weapons.

Despite failures in North Korean denuclearization campaigns over the past several decades, it is important to examine which steps had the potential to bring about the most effective changes and how those steps could be adapted to achieve long-term sustainability in the context of future negotiations.

It seems evident that halting measures, such as technical freezes and political moratoria, are likely to act as precursors to subsequent dismantlement steps, and if sequenced correctly, can themselves act as rollback measures due to the challenges of resuming operations after long periods of time. The disablement steps that North Korea took throughout the past three decades––in particular at Yongbyon––represented a proof-of-concept for how halting measures could be utilized to make it more difficult for North Korea to restart nuclear operations after a prolonged pause. Their relative isolation and short timeframes, however, made these steps easily reversible when the political climate worsened. This demonstrates that isolated halting measures are themselves highly fragile.

This article highlights the various factors that have contributed to the failure to move beyond halting measures over the past three decades of negotiations with North Korea. In particular, maximalist objectives, disagreements over sequencing, incompatible language, and the inability to take advantage of technical or political momentum presented insurmountable challenges for successive rounds of negotiations. Additional research is required in order to develop a roadmap for unblocking these systemic barriers; if momentum and sequencing can be sustained, then the compounded effects of these measures could begin to have more long-standing technical and economic impacts North Korea’s ability to reverse or restart its nuclear weapons program. It seems clear, however, that until these hurdles are surpassed, future halting measures are unlikely to yield sufficient progress to achieve ‘adequate irreversibility’.

Acknowledgments

The authors would like to thank the organizers of the Consortium study on irreversibility in nuclear disarmament, and particularly Hassan Elbahtimy. The authors would also like to thank their colleagues at the Federation of American Scientists’ Nuclear Information Project for their continued guidance.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by King’s College London.

Notes on contributors

Matt Korda

Matt Korda is a senior research fellow for the Nuclear Information Project at the Federation of American Scientists, and an associate researcher with the Nuclear Disarmament, Arms Control and Non-proliferation Programme at the Stockholm International Peace Research Institute (SIPRI). Previously, he worked for the Arms Control, Disarmament, and WMD Non-Proliferation Centre at NATO headquarters in Brussels. Korda’s research and open-source discoveries about nuclear weapons have made headlines across the globe, and his work is regularly used by governments, policymakers, academics, journalists, and the broader public in order to challenge assumptions and improve accountability about nuclear arsenals and trends. He received his MA in International Peace and Security from the Department of War Studies at King’s College London.

Eliana Johns

Eliana Johns is a research associate for the Nuclear Information Project at the Federation of American Scientists, where she researches the status and trends of global nuclear forces and the role of nuclear weapons in military doctrine. Previously, Johns worked as a project associate for DPRK Counterproliferation at CRDF Global, focusing on WMD nonproliferation initiatives to curb North Korea’s ability to gain revenue to build its weapons programs. Johns graduated with her bachelor’s in political science with minors in Music and Korean from the University of Maryland, Baltimore County (UMBC).